Identification of global histogram equalization by modeling gray-level cumulative distribution

Previous studies have shown how to “fingerprint” certain image processing operation. In this paper, we model the gray-level cumulative distribution of histogram-equalized image as discrete identity function, which is an intrinsic fingerprint produced by global histogram equalization processing. This model is shown to match well with the observed gray-level cumulative distribution. Based on the new model, a classification method is proposed for identifying the use of global histogram equalization processing. In comparison to previous arts, the proposed method can accurately distinguish global histogram equalization from other kinds of contrast enhancement techniques. The effectiveness of the proposed method is exhaustively evaluated in the context of histogram-equalized image identification and resistance against attacks.